Pharmacology, Biochemistry and Behavior 133 (2015) 18–24

Contents lists available at ScienceDirect

Pharmacology, Biochemistry and Behavior journal homepage: www.elsevier.com/locate/pharmbiochembeh

Cocaine decreases saccharin preference without altering sweet taste sensitivity Jennifer K. Roebber a, Sari Izenwasser a,b, Nirupa Chaudhari a,c,⁎ a b c

Graduate Program in Neurosciences, University of Miami Miller School of Medicine, Miami, FL 33136, USA Department of Psychiatry and Behavioral Sciences, University of Miami Miller School of Medicine, Miami, FL 33136, USA Department of Physiology and Biophysics, University of Miami Miller School of Medicine, Miami, FL 33136, USA

a r t i c l e

i n f o

Article history: Received 4 September 2014 Received in revised form 10 March 2015 Accepted 16 March 2015 Available online 24 March 2015 Keywords: Anticipatory contrast Cocaine Taste sensitivity Lickometer Brief access taste test Saccharin

a b s t r a c t In rodents, saccharin consumption is suppressed when the sweet taste stimulus is paired with moderate doses of cocaine. Several hypotheses have been used to explain the seemingly contradictory effect of decreased consumption of a normally preferred substance following a highly rewarding drug. A common theme across these hypotheses is that saccharin is interpreted as less rewarding after cocaine pairing. We considered the alternative possibility that suppression is caused not by a change in reward circuitry, but rather by a change in taste detection, for instance by altering the afferent taste response and decreasing sensitivity to sweet taste stimuli. To evaluate this possibility, we measured saccharin taste sensitivity of mice before and after a standard cocaine-pairing paradigm. We measured taste sensitivity using a brief-access lickometer equipped with multiple concentrations of saccharin solution and established concentration–response curves before and after saccharin– cocaine pairing. Our results indicate that the EC50 for saccharin was unaltered following pairing. Instead, the avidity of licking saccharin, an indicator of motivation, was depressed. Latency to first-lick, a negative indicator of motivation, was also dramatically increased. Thus, our findings are consistent with the interpretation that saccharin–cocaine pairing results in devaluing of the sweet taste reward. © 2015 Elsevier Inc. All rights reserved.

1. Introduction Cocaine exposure and consumption of sweet tasting solutions both cause increased dopamine signaling in the nucleus accumbens, leading to pleasurable or rewarding sensations (Avena et al., 2008; Mark et al., 1991; Rada et al., 2005). These two stimuli enhance dopamine signaling through separate mechanisms. Cocaine directly inhibits the dopamine transporter, thereby elevating synaptic dopamine levels. In contrast, sucrose elicits neural circuit activity that indirectly increases dopamine release in the nucleus accumbens (Avena et al., 2008; Norgren et al., 2006; Rada et al., 2005). There also are notable differences in the reward value and specific pathways for sweet stimuli and cocaine. For example, dopamine firing patterns in the nucleus accumbens are usually blunted after repeated access to sweet rewards, but not cocaine (Di Chiara et al., 2004; Rada et al., 2005). In addition, lesion studies have shown that taste circuits in the thalamus and projections from the parabrachial nucleus are essential for sweet- but not cocaine-elicited reward (Grigson et al., 2000; Norgren et al., 2006; Nyland et al., 2012). It has been known for many years that rodents will avoid intake of a palatable taste cue when it is repeatedly presented prior to an injection ⁎ Corresponding author at: Department of Physiology and Biophysics, 1600 NW 10th Ave., Miami, FL 33136, USA. Tel.: +1 305 243 3187; fax: +1 305 243 5931. E-mail address: [email protected] (N. Chaudhari).

http://dx.doi.org/10.1016/j.pbb.2015.03.010 0091-3057/© 2015 Elsevier Inc. All rights reserved.

of cocaine (Ferrari et al., 1991; Glowa et al., 1994; Grigson, 1997). This appears to cause a paradox where a reinforcing drug of abuse causes avoidance rather than enhanced pursuit of the conditioned stimuli. There are several behavioral hypotheses that seek to explain this observation. For instance, changes in homeostasis following drug injection may cause a ‘taste shyness’ around the conditioned tastant (Hunt and Amit, 1987). Alternatively, anticipation of the future high drug reward may dampen the value of the less rewarding natural stimulus (Grigson, 1997). Other hypotheses include the idea that the drug is aversive to the mice and induces conditioned taste aversion (Lin et al., 2014) or conditioned disgust (Parker et al., 2008). Another possibility is that there may be an afferent, taste-driven explanation for the decreased consumption. For example, decreased ability to taste the cue (saccharin) may result in decreased consumption following cocaine exposure and conditioning. Such a mechanism may be quite independent of hedonic value and motivation. It is known that cocaine inhibits several transporter proteins, thus blocking transport, not only of dopamine, but also of the other monoamines, serotonin and norepinephrine (Amara and Sonders, 1998; Mateo et al., 2004). Peripheral taste cells employ both serotonin and norepinephrine as neurotransmitters (Huang et al., 2005, 2007) and the transporters for norepinephrine and serotonin are present and functional on cells of taste buds (Chaudhari and Roper, 2010; Dvoryanchikov et al., 2007). It also has been shown that knockout of both dopamine and serotonin

J.K. Roebber et al. / Pharmacology, Biochemistry and Behavior 133 (2015) 18–24

transporters are required to abolish cocaine conditioned place preference, even though dopamine remains the primary neurotransmitter underlying the rewarding effects of cocaine (Sora et al., 2001). Similarly, mice with genetic deletion of transporters for serotonin or norepinephrine show altered responses to taste-selective behavior (Jones et al., 2010). Cocaine also has been shown to alter many aspects of appetite and ingestion (Cui and Lutter, 2013; Vicentic and Jones, 2007). Cocaine dependence in humans causes profound changes in diet, specifically by increasing consumption of fats and carbohydrates, but decreasing sugar intake (Ersche et al., 2013). If cocaine acutely alters signaling in taste buds, repeated exposures may well result in chronically altered gene expression. Cocaine has been shown to lead to long-term transcriptional alterations through CREB expression (Xu et al., 2013) and chromatin remodeling (Taniguchi et al., 2012). Cocaine could alter taste sensitivity by modifying gene expression in existing and particularly in newly formed taste bud cells, which are replaced every few days (Beidler and Smallman, 1965). In support of this, the time-frame for cocaine-induced is similar to that of taste bud cell turnover. Thus, we considered whether long-term exposure to cocaine might change chemosensory detection and neural signals from the taste periphery. The goal of the present study was to apply a model commonly used in taste studies but not commonly used in drug studies to measure rodent taste sensitivity before and after saccharin is paired with cocaine to determine if the pairing paradigm causes taste sensitivity to be dampened. We used a brief-access lickometer assay, as it is commonly used to identify taste quality and to separate orosensory-mediated responses from higher order responses (Loney et al., 2012; Mantella and Youngentob, 2014; McCaughey and Glendinning, 2013). This assay has been used to quantify differences of taste sensitivity in mice with mutations in taste-related genes (Tordoff and Ellis, 2013; Treesukosol et al., 2009) and to compare orosensory ability and drug preference for both nicotine (Glatt et al., 2009) and alcohol (Brasser et al., 2012). Brief-access lickometry has been optimized to elicit concentration-dependent licking for different taste stimuli and to effectively separate mice with altered taste function (Glendinning et al., 2002; Sinclair et al., 2014). A brief-access lickometer assay generates a concentration–response curve of the acceptance of a taste stimulus. Importantly, voluntary consumption, preference in a 2-bottle assay, and brief-access lick-rate all are dependent on similar concentrations of the stimulus and display parallel responses (Inoue et al., 2007). A substantial difference between the brief-access lickometer assay and the commonly used two-bottle preference test is that licking responses reflect taste-dependent acceptance or rejection that is minimally influenced by post-ingestive effects. Thus, we used a Davis rig lickometer to determine if detection of a sweet taste stimulus is altered when mice are exposed to a standard saccharin–cocaine pairing paradigm. We administered this brief-access test both before and after cocaine exposure (pre-test vs. post-test, respectively). 2. Materials and methods 2.1. Animals Sixteen male C57BL/6 mice between 2 and 4 months of age were individually housed on a 12-h light/dark cycle with food and water available. All mice were handled according the 2010 NIH guide for the Care and Use of Laboratory Animals, 8th ed. and all procedures were approved by the University of Miami Institutional Animal Care and Use Committee (IACUC). Male mice were used exclusively to match conditions used in previous studies. 2.2. Drugs and dosage Cocaine was obtained from NIDA (Rockville, MD, USA) and was dissolved in 0.9% NaCl. Cocaine (30 mg/kg) was injected i.p. for

19

experimental animals while an equal volume (0.01 ml/g) of normal saline was injected for controls. Dosage and procedures for pairing saccharin and cocaine in C57BL/6J mice were followed exactly as described in Freet et al. (2013). Saccharin solutions were made by dissolving saccharin sodium salt hydrate (Sigma #S1002) in deionized autoclaved water. 2.3. Water and Food deprivation Glendinning et al. (2002) devised a mild water- and food-restriction paradigm that motivates mice to lick in a Davis Rig Lickometer and optimizes the concentration-dependence of the response (Glendinning et al., 2002). Prior to Lickometer Training Sessions 1 and 2, mice were fully deprived of water for 23.5 h. This familiarizes the animals to the drinking spout. For all subsequent sessions, mice were subjected to a limited water and food restriction: 1 g LabDiet 5001 mouse chow plus 3 ml of water for the 23.5 h prior to testing (Glendinning et al., 2002). All animals were weighed and examined daily for weight loss or behavioral changes associated with dehydration. None of the animals showed a N20% weight drop or sudden increase in lethargy during food or water restriction. 2.4. Lickometer Sessions Each mouse was placed in a Davis MS-160 “lickometer” (DiLog Instruments Tallahassee, FL) during seven separate sessions. On the first training day, mice were water deprived for 23.5 h prior to the experiment and placed individually in the lickometer cage for 1 h to familiarize them with the novel environment. A stationary drinking spout was present and the mouse was free to drink water ad libitum. Mice were given 1 h following training to recover in their home cage with full access to food and water before being water deprived for the next session. On Training Day 2, the mouse was placed in the lickometer cage for 30 min and was introduced to a non-stationary drinking tube. During this training session, a bottle of water was presented and the mouse was allowed to drink for five seconds before a shutter prevented further drinking. After a 15 second wait, a new bottle of water was presented. Training Day 3 serves to familiarize mice with repeated presentations of varying concentrations of saccharin, representing a novel taste (Sinclair et al., 2014). On Training Day 3 and on all test days, seven bottles were filled with either water or a solution of sodium saccharin (0.9 mM, 2 mM, 7 mM, 9 mM, 18 mM, or 45 mM corresponding to 0.018%, 0.041%, 0.18%, 0.37%, 0.92% g/ml, respectively) and were presented in computer-generated random order. Concentrations of saccharin were chosen to span the preferred concentration range in C57BL/6 mice (Bachmanov et al., 2001; Fuller, 1974; Inoue et al., 2007). In all tests, water and saccharin solutions were presented one at a time over a total of 30 min. or for a maximum of 49 presentations, whichever came first. Each presentation comprised of the shutter opening to reveal a drinking spout, a variable latency before the mouse began drinking, and 5 s of drinking time. Mice were given one day of rest with full hydration after each testing session. Every mouse was tested for all concentrations of saccharin twice prior to cocaine (or saline) exposure (Pre-test 1,2) and twice post-cocaine (or saline) exposure (Post-test 1,2). The post-tests were conducted at least 27.5 h after the last cocaine injection. Thus, mice were never tested in the lickometer while under the influence of cocaine. The time course of experiments is outlined in Fig. 1. 2.5. Data analysis: maximal lick rate and the Standardized Lick Ratio For every lickometer presentation, we recorded the inter-lick interval (in msec), the total number of licks, the identity of the solution, and the latency interval before the mouse started to drink in each trial. First, we calculated the maximum lick rate for each mouse as

20

J.K. Roebber et al. / Pharmacology, Biochemistry and Behavior 133 (2015) 18–24

Fig. 1. Experimental timeline. C57BL6J mice were trained in the lickometer over 7 days. For each training session, water deprivation or food and water restriction began 23.5 h prior to training. Successive training sessions introduce a new variable to the mouse each day: first, the lickometer itself; second, intermittently presented sipper tube; and third, varied saccharin concentrations. Rest days were introduced to prevent excessive dehydration and to promote consolidation of learning. Next, during the pre-test, mice were tested for their individual taste sensitivity and motivation to drink saccharin-sweetened solutions. In the following conditioning period of 10 days, mice were presented with saccharin (7 mM) for 1 h every other day, immediately followed by an injection of either cocaine (cocaine-injected mice) or saline (control mice). In the days immediately following conditioning (post-test), the mice were again tested for their taste sensitivity and motivation to drink saccharin solutions.

previously reported (Glendinning et al., 2002). In brief, the maximum rate for licking water was measured during Training Days 1 and 2, when water was used as a stimulus, and the mouse was maximally dehydrated. This value was used as the mouse's intrinsic maximum rate of oromotor function and motivation to drink. In the subsequent test days, mice were food and water restricted, but not deprived, and were expected to lick at saccharin solutions in a concentrationdependent fashion. The lick rate for each mouse to a particular solution was averaged across all presentations for that concentration on that test day, and was then normalized to the individual maximum lick rate for that mouse to calculate the Standardized Lick Ratio (SLR) (Glendinning et al., 2002). By normalizing mice to their own respective maximum rate, we controlled for inherent differences between oromotor function across individual mice. For example, if a mouse showed 10 licks/5 s (averaged across 7 presentations on Test Day 1) in response to 9 mM saccharin and its maximal lick rate was 40 licks/5 s, the resulting SLR would be 0.25. Once normalized, SLR was averaged across two days of testing for each concentration and all mice in the group (e.g. all 8 control mice were averaged across both days of pre-testing) (Fig. 3, data points). Averaged SLR values were then fit in Graphpad Prism with fourparameter dose-response curves (Fig. 3, curves). SLR ¼ Vmin þ ðVmax − Vmin Þ=ð1 þ 10ˆ½ð logEC50 −concentrationÞ  HillslopeÞ The equation above was used to model upper and lower asymptotes (Vmax and Vmin) and to determine the EC50, defined as the concentration that evokes a response half way between Vmax and Vmin. In this analysis, Vmax is the saturation or maximal lick response to saccharin. Vmin is modeled by the equation, rather than set to zero, because the mice must lick at least once to perform a trial. The fourth parameter that the four-parameter dose response models is the Hill Slope. Following cocaine exposure, we expected one of two changes to occur in the dose-response curve. If afferent taste sensitivity was decreased following cocaine exposure, we would expect to see a lateral shift to the right, as the mouse fails to differentiate low concentrations of saccharin from water, but can still taste very high concentrations of saccharin. This kind of behavioral output from a lickometer has been illustrated following transection of a subset of the nerves that carry taste responses (St John and Boughter, 2004). Alternatively, a change in Vmax, signifying decreased avidity would suggest that taste sensitivity

is unaltered, but that the hedonic value of the stimulus has changed (Ahmed and Koob, 1998). 2.6. Saccharin–cocaine pairing Saccharin–cocaine or saccharin–saline pairings were carried out for 10 days as described by Freet et al. (2013) and diagrammed in Fig. 1. Mice were placed on a water-restricted schedule where they were given access to fluids for 3 h daily. On alternating days, mice were given either water or 7 mM saccharin for 1 h in the morning (10– 11 AM). Immediately following the one-hour access to saccharin, half of the mice were injected intraperitoneally (i.p.) with cocaine (30 mg/kg), while the remaining mice were injected with an equal volume of saline (all injections were administered in a volume of 0.01 ml/g body weight). Water was additionally provided ad libitum for 2 h every afternoon (3–5 PM) to prevent excessive dehydration. After the final injection, mice were given full access to food and water for 3 h and then were subjected to the food and water restriction schedule for lickometer testing. Mice were weighed and examined prior to each session for signs of dehydration. 2.7. Statistical analysis Statistical analyses were conducted and graphs were produced in Graphpad Prism v5. Data are presented for seven control mice and eight cocaine-exposed mice. One of the control mice was removed before post-tests due to poor health. Data from this mouse was retained in pairing data (Fig. 2). However, since the mouse did not complete concentration–response post-tests, the corresponding data were removed from all concentration–response tests (Figs. 3 and 4). Averaged standard lick ratios (SLRs) for all concentrations were analyzed using a 2-way ANOVA. Averaged SLRs were then fit to create nonlinear fourparameter dose response curves. These curves modeled maximal and minimal responses for liking saccharin (Vmax and Vmin), as well as the EC50 and Hill Slope. To confirm curve fit trends, we analyzed individual changes in mice by running paired two-tailed Student's t-tests for the SLR of the maximum concentration tested as well as the latency to drink. 3. Results 3.1. Pre-test Taste sensitivity for saccharin was assessed by brief access lickometer tests using a range of saccharin concentrations. We

J.K. Roebber et al. / Pharmacology, Biochemistry and Behavior 133 (2015) 18–24

Fig. 2. Consumption of saccharin solution and water during the saccharin–cocaine pairing interval. Data are means ± SEM of intake volumes. A, Intake of 7 mM saccharin solution, by mice, in the 1 h prior to injection of saline (○) or cocaine (•). B, Intake of water by the same groups of mice as in A and at the same time of day, but on alternating days with no injections and no access to saccharin. Saccharin and water consumption occurred on alternating days in the home cage. Cocaine–saccharin pairing occurred on days marked by (↑).

performed identical tests before and after exposure to cocaine. All lickrate measurements were normalized to each mouse's intrinsic maximum rate of licking, eliminating much animal-to-animal variability. Saccharin concentration–response curves prior to cocaine exposure established baseline sensitivity. The profiles we obtained are very similar to earlier measurements on C57BL6 mice (Inoue et al., 2007). We also used these curves to ensure that the experimental and control groups of mice were not significantly different at the start of the experiment. The EC50 and Vmax values of each concentration–response curve were obtained following non-linear fits to the data. The EC50 for saccharin at the beginning of the protocol was 10.4 mM for the cocaine group and 13.0 mM for the control group. Before exposure, neither the EC50 nor the Vmax was significantly different across the two groups (Table 1). Additionally, the two groups of mice did not differ significantly in the time between presentation of a tube and initiation of drinking, referred to here as latency (15.01 s and 16.54 for cocaine and control groups respectively t(13) = 0.520, p = 0.510; two-tailed Student's t-test).

21

Fig. 3. Cocaine alters the avidity of licking saccharin solutions without changing the EC50. A, Concentration–response curves for saccharin (n = 7 mice) before (○) and after (▼) saline injections paired with saccharin. B, Saccharin concentration–response curves (n = 8 mice) before and after cocaine injections paired with saccharin. Data points on graph represent means ± SEM of SLR values calculated as described (Materials and methods). Curves represent 4-parameter non-linear fits to the data (R2 for pre- and post-tests in saline-injected mice, 0.998, 0.999; and in cocaine-injected mice, 0.996, 0.871 respectively). There was no significant change in calculated EC50 (95% CI) for either control or cocaine exposed mice. On the other hand, there was a significant change in Vmax (95% CI) in cocaine-exposed mice, but not in saline-exposed mice compared to the respective pre-tests. We also analyzed the individual SLR data points across all saccharin concentrations by 2-way ANOVA. *Mice exhibited significantly lower SLR values in the post-cocaine test (2-way ANOVA of saccharin concentration vs. drug, F(1) = 27.40, p b 0.0001) whereas control mice were not altered (F(1) = 2.14, p = 0.148) following injections.

previous studies (Freet et al., 2013). Control mice that were injected with saline showed no decline in saccharin consumption over time (Fig. 2A). The two groups of mice showed no differences in consumption of the unconditioned stimulus (water, Fig. 2B), which was presented every other day in place of saccharin. Neither group showed changes in water consumption as a function of day, consistently drinking ≈ 1.7 ml of water (Fig. 2B). 3.3. Post-test

3.2. Saccharin–cocaine pairing We paired saccharin consumption with exposure to cocaine by first allowing mice to drink a saccharin solution (7 mM) ad libitum for 1 h and then injecting them i.p. with cocaine. This procedure was repeated over multiple 48-hour trials. A control group of mice was treated identically except that they were injected with saline. The volume of saccharin consumed was measured and we found that the cocaine group decreased their saccharin consumption relative to their initial rate after their second injection of cocaine (Fig. 2A), consistent with

Following the 10-day period of cocaine exposure, we re-assessed the sensitivity and motivation of mice to drink saccharin solutions using the same brief-access lickometer test as before. Control (saline-injected) mice displayed unaltered concentration–response curves for saccharin. For control mice, a 2-way ANOVA (drug × concentration) on SLR values of each concentration of saccharin showed that there was no significant difference between the pre- and post-tests, F(1) = 2.14, p = 0.148. This demonstrated that the mice retained lickometer training over the 10day interval of injections. The 2-way ANOVA further showed that

22

J.K. Roebber et al. / Pharmacology, Biochemistry and Behavior 133 (2015) 18–24

To confirm that the response to the highest saccharin concentration was indeed altered for each individual mouse following cocaine injections, we performed within subject (paired) two-tailed Student's t-tests. The SLR for each mouse at the highest concentration (45 mM) of saccharin was compared in the pre-test and post-test. In salineinjected control mice, the SLR to 45 mM saccharin was unchanged following treatment (t(6) = 0.902, p = 0.40). In contrast, cocaineinjected mice showed a significantly lower SLR to 45 mM saccharin after the cocaine-treatment (t(7) = 2.5, p = 0.04). That is, while control mice licked the highest concentration of saccharin at 106% of their pre-test rate (an insignificant change), the cocaine-exposed mice dramatically decreased their licking to 69% of pre-test value. 3.4. Latency

Fig. 4. Latency dramatically increases after saccharin is paired with cocaine injections but not after saline injections. Latency is the average time between when a saccharin solution is presented and when the mouse starts to drink. The mice undergo 10 days of saccharin– cocaine or saccharin–saline pairing between the Pre-tests and Post-tests. Data are means ± SEM of latencies observed across two test days. *For the cocaine mice, latency during post-tests was significantly longer than during pre-tests (paired 2-tailed Student's t-tests, t (13) = 3.1, p = 0.007). Control mice did not show any difference (t (15) = 1.9, p = 0.08).

there was not a significant difference F(1) = 0.75, p = 0.590 between pre- and post-tests for 7 mM saccharin, the conditioning concentration used during the intervening 10-days. That is, control mice did not alter their response to the conditioning concentration of saccharin. Mice in which saccharin was paired with cocaine showed a significant change in their dose-response curves in the post-test. A 2-way ANOVA (drug × saccharin concentration) on SLR values demonstrated that there was a highly significant difference, (F(1) = 27.40, p b 0.0001), between the pre- and post-tests. That is, dose response curves were significantly changed by cocaine pairing. Again, no interaction was found between saccharin concentration and drug (F(5) = 1.54, p = 0.186), showing that alteration of licking behavior in cocaine treated mice was generalized to all concentrations of saccharin, not only the particular concentration (7 mM) that had previously been paired with cocaine. 4-parameter curve fits of lickometer data (Fig. 3) showed that cocaine-exposed mice had a drop in Vmax in the concentration–response curve for saccharin but no shift in EC50 (Fig. 3B, Table 1). This shows that cocaine pairing caused a curve compression without any lateral shifting. The post-exposure Vmax for saccharin for the two groups, Control and Cocaine were 0.927 and 0.529 respectively (Table 1). These calculated values are significantly different across the groups (95% confidence interval).

During lickometer testing, latency is defined as the time between when a tube is presented and when the mouse starts to drink. A short latency results when the mouse initiates drinking soon after the sipper tube is presented. Since presentations are short and frequent (up to forty-nine presentations, each lasting no more than five seconds) and randomized, a consistently short latency indicates both a familiarity with the task and a willingness to drink low and high concentrations of saccharin. During the pre-test, the short latency is a good indication of task learning and interest in the saccharin. Following pairing of saline and saccharin, the control mice did not show a significant change in latency. In contrast, the mice in which saccharin was paired with cocaine exhibited a dramatic increase in latency (Fig. 4). Two-tailed paired Student's t-tests confirmed that mice had significantly increased latencies to licking saccharin after cocaine (t(15) = 3.11, p = 0.007), whereas control mice showed no significant change in latency (t(13) = 1.90, p = 0.08). That is, every cocaine-injected mouse approached the drinking spout with a greater delay after the 10-day interval of exposure to cocaine. In a Bonferroni multiple comparison test, we confirmed that pre-test latencies across the two groups were not significantly different (p N 0.1), nor were pre- vs. post-test latencies different for the control group (p N 0.1). However, within the cocaine group, pre- and post-test latencies differed significantly (p b 0.001). Finally, there were significant differences in post-test latencies (p b 0.001) when comparing the control group to the cocaine group. Our data demonstrated that both parameters, depression of maximum licking response and increase of latency, were altered following the saccharin–cocaine pairing. Indeed, we found a significant linear correlation (R2 = 0.62) between the two parameters as measured for each mouse: increase in latency and depression of lick rate at 45 mM saccharin (the highest concentration tested). The data suggest that both decreased avidity of licking and increased latency (i.e. decreased anticipation) result from the saccharin–cocaine pairing paradigm. 4. Discussion

Table 1 Parameters calculated for saccharin concentration–response curves in the two groups of mice. Standardized Lick Ratio (SLR) data were fit to sigmoidal concentration–response curves to derive values for EC50 and Vmax for the control (n = 7) and cocaine (n = 8) group. Values presented are best fit for each parameter ± 95% CI. Neither the control group nor the cocaine group showed a significant difference in the derived EC50 values for their respective pre- and post-tests. EC50 values did not significantly change from pre-test to post-test within or across the groups.

EC50 (mM) VMax (SLR)

Control group (n = 7)

Cocaine group (n = 8)

Pre-test

Post-test

Pre-test

Post-test

13.0 ± 4.16 0.877 ± 0.165

15.0 ± 2.83 0.927 ± 0.110+

10.4 ± 2.97 0.767 ± 0.128

12.6 ± 4.24 0.529 ± 0.105⁎

+ The calculated Vmax for the post-tests was significantly different between the control and cocaine groups. ⁎ The calculated Vmax for the cocaine group was significantly different between pre- and post-tests.

4.1. Depression of response to saccharin Previous studies have shown that after repeatedly pairing cocaine with a sweet liquid like saccharin, the consumption of saccharin is decreased (Grigson, 1997; Parker, 1995; Risinger and Boyce, 2002). This phenomenon has been interpreted a number of different ways. One hypothesis is that an anticipatory contrast effect develops where anticipation of the greater reward of cocaine decreases the normal reward value of saccharin (Grigson, 1997; Nyland et al., 2012). A competing hypothesis states that a form of taste avoidance learning occurs, where the drug-induced disruption of homeostasis causes a ‘taste shyness’ around saccharin when not self-administered (Hunt and Amit, 1987). There are additional aversive properties of foods and drugs (Riley, 2011), suggesting that conditioned taste aversion (Lin et al., 2014) or conditioned disgust (Parker et al., 2008) could be at

J.K. Roebber et al. / Pharmacology, Biochemistry and Behavior 133 (2015) 18–24

play. In these cases, it is hypothesized that the nauseating effects of the drug overpower the normal reward of the drug and its associated conditioned stimulus. However, this last explanation is confounded by studies that show mechanistic differences between LiCl-induced conditioned taste aversion and the avoidance of saccharin when paired with drugs (Grigson and Freet, 2000; Grigson et al., 2000). While a variety of behavioral models exist to explain this phenomenon, the present study aimed to answer a simpler question. Could cocaine exposure, through a standard saccharin–cocaine pairing paradigm, cause a decrease in taste sensitivity? Thus, our goal was to apply a model commonly used to quantify taste responses to the wellstudied phenomenon of drug-induced suppression of saccharin preference. In brief-access lickometry, taste stimuli are presented for a short (seconds) duration. The output lick rate is not directly related to consumption in calories or volume, but instead, is a function of orosensory detection with minimal influence from post-ingestive effects (Glendinning et al., 2002; Loney et al., 2012; McCaughey and Glendinning, 2013). 4.2. Saccharin concentration response curves By measuring brief access licking responses to multiple varying concentrations of saccharin, we have shown that mice exhibit a consistent concentration–response behavior towards saccharin. After pairing cocaine exposure with saccharin drinking, we found no significant change in EC50 for saccharin. These data show that despite the potential mechanisms available for modification of taste detection, cocaine does not produce a lasting change in taste sensitivity under these conditions. A vertical shift of the licking concentration–response curve would suggest a change in hedonic set point (Ahmed and Koob, 1998), whereas a lateral shift on the concentration axis would suggest tolerance (right shift) or sensitization (left shift) to the taste solution. Aversive stimuli such as quinine (bitter) produce a negatively oriented doseresponse curve with decreasing lick rates to progressively higher stimulus concentrations (St John and Boughter, 2004). When a stimulus becomes aversive as a consequence of conditioning, the licking response is transformed from a preferred (high lick-rate) to an aversive (suppressed licking) pattern (Hashimoto and Spector, 2014). We found that after pairing saccharin with cocaine, the concentration response curve displayed a dramatic vertical depression with no lateral shift or sign-change (Fig. 3B). That is, animals experienced a depression in the assigned value of the taste stimulus, without a change in EC50. Because brief-access lickometer behavior is driven by taste detection, the unchanged EC50 suggests cocaine exposure did not change the ability of mice to taste saccharin. Instead, our experimental data suggest a decreased palatability or preference for saccharin after this previously preferred stimulus is associated with the drug. For saccharin and other non-caloric preferred stimuli, the short-term lick-rate in the briefaccess lickometer assay is taste-dependent and parallels long-term consumption (Inoue et al., 2007). Thus, our findings of depressed licking are consistent with previously proposed hypotheses (Grigson, 1997; Hunt and Amit, 1987) that decreased saccharin consumption after cocaine exposure is caused by diminished palatability of saccharin, either through a change in relative value (anticipatory contrast) or a mild avoidance or shyness around the drug. However, in our study, mice retained increased lick rates for higher concentrations of saccharin following saccharin–cocaine pairings. Thus, our findings are inconsistent with the premise of a conditioned disgust or strong conditioned aversion (Lin et al., 2014) (Parker et al., 2008), which would result in an inverted dose-response curve, with decreased licking of increasingly concentrated saccharin. We were mindful that chronic food restriction has been shown to dampen the contrast between more and less rewarding foods (Cottone et al., 2009). We believe this phenomenon likely did not influence behavior in our experiments. First, the paradigm we used, established by (Glendinning et al., 2002) is a mild restriction for a

23

short duration, and results in enhanced contrast between water and saccharin (Glendinning et al., 2002). Second, we ensured that food restriction occurred for only 1 day immediately prior to lickometer sessions, and did not occur during the 10-day interval of cocaine exposure. 4.3. Latency In the present study, not only did mice lick saccharin solution with less avidity after exposure to cocaine, but they also took longer to initiate drinking during each trial. The literature indicates that increased latency is not associated with unpalatability or with decreased taste sensitivity. In particular, latency increases do not accompany increasing concentrations of aversive (bitter) taste stimuli; nor are they a hallmark of decreased sensory input such as transection of taste nerves (St John and Boughter, 2004). Latency to initiate licking has been studied in several contexts. A short latency is generally associated with oral acceptability (Mantella and Youngentob, 2014; Nyland et al., 2012), while a long latency can be indicative of failure to learn the task, or of feared or unpleasant stimuli (Houpt et al., 2007). In our experiments, the mice are unaware of the precise stimulus presented. Hence, altered latency to approach the stimulus may not reflect a specific response to the quality or concentration of saccharin. When exposed to ethanol, mice increasingly seek out natural rewards. This is seen in lickometer studies as increased licking and decreased latency to initiate sucrose trials (Pastor et al., 2010). Cocaine treatments, on the other hand, have the opposite effect. Cocaine impairs motivation for demanding tasks. This is seen in a lowered break point in the number of lever presses to acquire a sucrose reward, as well as decreased motivation to mate with a receptive female (Barnea-Ygael et al., 2014). Considering previous literature, our observation of increased latency to lick saccharin solutions is concordant with the known effect of long-term cocaine exposure. Although our mice still found the saccharin appetitive, as evidenced by increased licking of higher concentrations, it would seem that they are less motivated to drink saccharin. Further, our finding that the depression of lick rate is linearly correlated with increased latency adds further weight to the notion that the increase in latency is not due to gustatory interpretation of a less tasty stimulus, but rather a motivational change that counters normal saccharin reward. We noted, however, that both parameters were more strongly influenced by cocaine pairing in some mice than in others. 5. Conclusions Our findings reveal that following a standardized 10-day pairing of cocaine and saccharin, sensitivity for saccharin (measured as EC50) remains unaltered. Our data indicate that despite several potential mechanisms available for taste modification, cocaine exposure does not change taste sensitivity. Rather, following saccharin–cocaine pairing, decreased licking of saccharin along with increased latency to drink suggests a decreased motivation for a previously rewarding sweet stimulus. Our findings add further support to prior interpretations that the decrease in saccharin consumption likely is due to a change in the value of saccharin as a reward. Acknowledgments Supported through NIH grant R01 DC006308 (N.C.). References Ahmed SH, Koob GF. Transition from moderate to excessive drug intake: change in hedonic set point. Science 1998;282:298–300.

24

J.K. Roebber et al. / Pharmacology, Biochemistry and Behavior 133 (2015) 18–24

Amara SG, Sonders MS. Neurotransmitter transporters as molecular targets for addictive drugs. Drug Alcohol Depend 1998;51:87–96. Avena NM, Rada P, Hoebel BG. Evidence for sugar addiction: behavioral and neurochemical effects of intermittent, excessive sugar intake. Neurosci Biobehav Rev 2008;32: 20–39. Bachmanov AA, Tordoff MG, Beauchamp GK. Sweetener preference of C57BL/6ByJ and 129P3/J mice. Chem Senses 2001;26:905–13. Barnea-Ygael N, Gal R, Zangen A. Chronic cocaine administration induces long-term impairment in the drive to obtain natural reinforcers in high- but not lowdemanding tasks. Addict Biol 2014, Nov 12. http://dx.doi.org/10.1111/adb.12196. Beidler LM, Smallman RL. Renewal of cells within taste buds. J Cell Biol 1965;27: 263–72. Brasser SM, Silbaugh BC, Ketchum MJ, Olney JJ, Lemon CH. Chemosensory responsiveness to ethanol and its individual sensory components in alcohol-preferring, alcoholnonpreferring and genetically heterogeneous rats. Addict Biol 2012;17:423–36. Chaudhari N, Roper SD. The cell biology of taste. J Cell Biol 2010;190:285–96. Cottone P, Sabino V, Steardo L, Zorrilla EP. Consummatory, anxiety-related and metabolic adaptations in female rats with alternating access to preferred food. Psychoneuroendocrinology 2009;34:38–49. Cui H, Lutter M. The expression of MC4Rs in D1R neurons regulates food intake and locomotor sensitization to cocaine. Genes Brain Behav 2013;12:658–65. Di Chiara G, Bassareo V, Fenu S, De Luca MA, Spina L, Cadoni C, et al. Dopamine and drug addiction: the nucleus accumbens shell connection. Neuropharmacology 2004; 47(Suppl. 1):227–41. Dvoryanchikov G, Tomchik SM, Chaudhari N. Biogenic amine synthesis and uptake in rodent taste buds. J Comp Neurol 2007;505:302–13. Ersche KD, Stochl J, Woodward JM, Fletcher PC. The skinny on cocaine: insights into eating behavior and body weight in cocaine-dependent men. Appetite 2013;71:75–80. Ferrari CM, O'Connor DA, Riley AL. Cocaine-induced taste aversions: effect of route of administration. Pharmacol Biochem Behav 1991;38:267–71. Freet CS, Arndt A, Grigson PS. Compared with DBA/2J mice, C57BL/6J mice demonstrate greater preference for saccharin and less avoidance of a cocaine-paired saccharin cue. Behav Neurosci 2013;127:474–84. Fuller JL. Single-locus control of saccharin preference in mice. J Hered 1974;65:33–6. Glatt AR, Denton K, Boughter Jr JD. Variation in nicotine consumption in inbred mice is not linked to orosensory ability. Chem Senses 2009;34:27–35. Glendinning JI, Gresack J, Spector AC. A high-throughput screening procedure for identifying mice with aberrant taste and oromotor function. Chem Senses 2002;27:461–74. Glowa JR, Shaw AE, Riley AL. Cocaine-induced conditioned taste aversions: comparisons between effects in LEW/N and F344/N rat strains. Psychopharmacology 1994;114: 229–32. Grigson PS. Conditioned taste aversions and drugs of abuse: a reinterpretation. Behav Neurosci 1997;111:129–36. Grigson PS, Freet CS. The suppressive effects of sucrose and cocaine, but not lithium chloride, are greater in Lewis than in Fischer rats: evidence for the reward comparison hypothesis. Behav Neurosci 2000;114:353–63. Grigson PS, Lyuboslavsky P, Tanase D. Bilateral lesions of the gustatory thalamus disrupt morphine- but not LiCl-induced intake suppression in rats: evidence against the conditioned taste aversion hypothesis. Brain Res 2000;858:327–37. Hashimoto K, Spector AC. Extensive lesions in the gustatory cortex in the rat do not disrupt the retention of a presurgically conditioned taste aversion and do not impair unconditioned concentration-dependent licking of sucrose and quinine. Chem Senses 2014;39:57–71. Houpt TA, Cassell JA, Riccardi C, DenBleyker MD, Hood A, Smith JC. Rats avoid high magnetic fields: dependence on an intact vestibular system. Physiol Behav 2007; 92:741–7. Huang YJ, Maruyama Y, Lu KS, Pereira E, Plonsky I, Baur JE, et al. Mouse taste buds use serotonin as a neurotransmitter. J Neurosci Off J Soc Neurosci 2005;25:843–7. Huang YJ, Maruyama Y, Dvoryanchikov G, Pereira E, Chaudhari N, Roper SD. The role of pannexin 1 hemichannels in ATP release and cell–cell communication in mouse taste buds. Proc Natl Acad Sci U S A 2007;104:6436–41.

Hunt T, Amit Z. Conditioned taste aversion induced by self-administered drugs: paradox revisited. Neurosci Biobehav Rev 1987;11:107–30. Inoue M, Glendinning JI, Theodorides ML, Harkness S, Li X, Bosak N, et al. Allelic variation of the Tas1r3 taste receptor gene selectively affects taste responses to sweeteners: evidence from 129.B6-Tas1r3 congenic mice. Physiol Genomics 2007;32:82–94. Jones JD, Hall FS, Uhl GR, Riley AL. Dopamine, norepinephrine and serotonin transporter gene deletions differentially alter cocaine-induced taste aversion. Pharmacol Biochem Behav 2010;94:580–7. Lin JY, Arthurs J, Reilly S. Conditioned taste aversion, drugs of abuse and palatability. Neurosci Biobehav Rev 2014;45:28–45. Loney GC, Blonde GD, Eckel LA, Spector AC. Determinants of taste preference and acceptability: quality versus hedonics. J Neurosci Off J Soc Neurosci 2012;32:10086–92. Mantella NM, Youngentob SL. Prenatal alcohol exposure increases postnatal acceptability of nicotine odor and taste in adolescent rats. PLoS One 2014;9:e102255. Mark GP, Blander DS, Hoebel BG. A conditioned stimulus decreases extracellular dopamine in the nucleus accumbens after the development of a learned taste aversion. Brain Res 1991;551:308–10. Mateo Y, Budygin EA, John CE, Jones SR. Role of serotonin in cocaine effects in mice with reduced dopamine transporter function. Proc Natl Acad Sci U S A 2004;101:372–7. McCaughey SA, Glendinning JI. Experience with sugar modifies behavioral but not tasteevoked medullary responses to sweeteners in mice. Chem Senses 2013;38:793–802. Norgren R, Hajnal A, Mungarndee SS. Gustatory reward and the nucleus accumbens. Physiol Behav 2006;89:531–5. Nyland JE, Alexander DN, Liang NC, Grigson PS. Bilateral lesions of the thalamic trigeminal orosensory area dissociate natural from drug reward in contrast paradigms. Behav Neurosci 2012;126:538–50. Parker LA. Rewarding drugs produce taste avoidance, but not taste aversion. Neurosci Biobehav Rev 1995;19:143–57. Parker LA, Rana SA, Limebeer CL. Conditioned nausea in rats: assessment by conditioned disgust reactions, rather than conditioned taste avoidance. Can J Exp Psychol Rev Can Psychol Exp 2008;62:198–209. Pastor R, Kamens HM, McKinnon CS, Ford MM, Phillips TJ. Repeated ethanol administration modifies the temporal structure of sucrose intake patterns in mice: effects associated with behavioral sensitization. Addict Biol 2010;15:324–35. Rada P, Avena NM, Hoebel BG. Daily bingeing on sugar repeatedly releases dopamine in the accumbens shell. Neuroscience 2005;134:737–44. Riley AL. The paradox of drug taking: the role of the aversive effects of drugs. Physiol Behav 2011;103:69–78. Risinger FO, Boyce JM. Conditioning tastant and the acquisition of conditioned taste avoidance to drugs of abuse in DBA/2J mice. Psychopharmacology 2002;160:225–32. Sinclair MS, Perea-Martinez I, Abouyared M, St John SJ, Chaudhari N. Oxytocin decreases sweet taste sensitivity in mice. Physiol Behav 2015;141:103–10. Sora I, Hall FS, Andrews AM, Itokawa M, Li XF, Wei HB, et al. Molecular mechanisms of cocaine reward: combined dopamine and serotonin transporter knockouts eliminate cocaine place preference. Proc Natl Acad Sci U S A 2001;98:5300–5. St John SJ, Boughter Jr JD. The contribution of taste bud populations to bitter avoidance in mouse strains differentially sensitive to sucrose octa-acetate and quinine. Chem Senses 2004;29:775–87. Taniguchi M, Carreira MB, Smith LN, Zirlin BC, Neve RL, Cowan CW. Histone deacetylase 5 limits cocaine reward through cAMP-induced nuclear import. Neuron 2012;73: 108–20. Tordoff MG, Ellis HT. Taste dysfunction in BTBR mice due to a mutation of Itpr3, the inositol triphosphate receptor 3 gene. Physiol Genomics 2013;45:834–55. Treesukosol Y, Blonde GD, Spector AC. T1R2 and T1R3 subunits are individually unnecessary for normal affective licking responses to Polycose: implications for saccharide taste receptors in mice. Am J Physiol Regul Integr Comp Physiol 2009;296:R855–65. Vicentic A, Jones DC. The CART (cocaine- and amphetamine-regulated transcript) system in appetite and drug addiction. J Pharmacol Exp Ther 2007;320:499–506. Xu R, Serritella AV, Sen T, Farook JM, Sedlak TW, Baraban J, et al. Behavioral effects of cocaine mediated by nitric oxide-GAPDH transcriptional signaling. Neuron 2013;78: 623–30.